基于自组织数据挖掘的铁路客运量预测方法研究

Research on Forecast Method of Railway Passenger Transport Volume based on Group Method of Data Handling

由于传统的预测方法难以对影响铁路客运量变化的因素进行全面考虑,其预测精度不高。选择影响铁路客运量变化的因素:经济社会发展的原生性需求、铁路自身供给能力、不同交通方式、客运价格和旅行费用、运输服务质量等,建立基于自组织数据挖掘的铁路客运量预测模型。通过算例进行验算结果表明,自组织数据挖掘建模预测方法在变量多、数据少、普通的建模预测方法难以胜任建模任务的情况下,可以得到较满意的结果,适宜进行多因素的铁路客运量预测。

For traditional forecast model is difficult to carry overall consideration on factors influencing the change of railway passenger transport volume,so it has low forecast precision.The factors influencing the change of railway passenger transport volume should be selected,which including original demand of economic social development,railway self-supply capacity,different traffic mode,passenger transport cost and traveling cost and transport service quality,and the forecast model of railway passenger transport volume based on group method of data handling（GMDH） should be established.Through calculation by example,the result shows that,under the condition of common model establishment forecast method with many variables and less data is difficult to establish model,the forecast method based on GMDH could achieve satisfied result and suitable to take the forecast of railway passenger transport volume with multi-factors.